线性回归输出是一个连续值,因此适用于回归问题。回归问题在实际中很常见,如预测房屋价格、气温、销售额等连续值的问题。与回归问题不同,分类问题中模型的最终输出是一个离散值。我们所说的图像分类、垃圾邮件识别、疾病检测等输出为离散值的问题都属于分类问题的范畴。softmax回归则适用于分类问题。
接下来利用NDArray和autograd来实现一个线性回归的训练。首先,导入本节中实验所需的包或模块,其中的matplotlib包可用于作图,且设置成嵌入显示。
%matplotlib inline
from IPython import display
from matplotlib import pyplot as plt
from mxnet import autograd, nd
import random
我们构造一个简单的人工训练数据集,它可以使我们能够直观比较学到的参数和真实的模型参数的区别。设训练数据集样本数为1000,输入个数(特征数)为2。给定随机生成的批量样本特征 X∈R1000×2 ,我们使用线性回归模型真实权重 w=[2,−3.4]⊤ 和偏差 b=4.2 ,以及一个随机噪声项 ϵ 来生成标签y=Xw+b+ϵ, 其中噪声项 ϵ 服从均值为0、标准差为0.01的正态分布。噪声代表了数据集中无意义的干扰。下面,让我们生成数据集。
# set input feature number
num_inputs = 2
# set example number
num_examples = 1000
# set true weight and bias in order to generate corresponded label
true_w = [2, -3.4]
true_b = 4.2
features = torch.randn(num_examples, num_inputs,
dtype=torch.float32)
labels = true_w[0] * features[:, 0] + true_w[1] * features[:, 1] + true_b
labels += torch.tensor(np.random.normal(0, 0.01, size=labels.size()),
dtype=torch.float32)
plt.scatter(features[:, 1].numpy(), labels.numpy(), 1);
使用图像展示生成的数据集
读取数据集
def data_iter(batch_size, features, labels):
num_examples = len(features)
indices = list(range(num_examples))
random.shuffle(indices) # random read 10 samples
for i in range(0, num_examples, batch_size):
j = torch.LongTensor(indices[i: min(i + batch_size, num_examples)]) # the last time may be not enough for a whole batch
yield features.index_select(0, j), labels.index_select(0, j)
batch_size = 10
for X, y in data_iter(batch_size, features, labels):
print(X, '\n', y)
break
运行结果
tensor([[-0.2418, 0.0977],
[-1.4690, 1.0187],
[-0.8680, 1.3924],
[-0.2509, 0.0042],
[-0.8196, 2.7727],
[-1.3803, 1.3686],
[ 0.3174, -0.5171],
[-0.7607, 1.2662],
[ 0.6477, -0.4813],
[ 0.4225, -1.1675]])
tensor([ 3.3814, -2.1989, -2.2673, 3.6702, -6.8637, -3.2140, 6.5941, -1.6277,
7.1394, 9.0192])
初始化模型参数
w = torch.tensor(np.random.normal(0, 0.01, (num_inputs, 1)), dtype=torch.float32)
b = torch.zeros(1, dtype=torch.float32)
w.requires_grad_(requires_grad=True)
b.requires_grad_(requires_grad=True)
定义用来训练参数的训练模型
def linreg(X, w, b):
return torch.mm(X, w) + b
定义损失函数(均方误差函数)
def squared_loss(y_hat, y):
return (y_hat - y.view(y_hat.size())) ** 2 / 2
定义优化函数(小批量随机梯度下降)
def sgd(params, lr, batch_size):
for param in params:
param.data -= lr * param.grad / batch_size # ues .data to operate param without gradient tra
模型训练
# super parameters init
lr = 0.03
num_epochs = 5
net = linreg
loss = squared_loss
# training
for epoch in range(num_epochs): # training repeats num_epochs times
# in each epoch, all the samples in dataset will be used once
# X is the feature and y is the label of a batch sample
for X, y in data_iter(batch_size, features, labels):
l = loss(net(X, w, b), y).sum()
# calculate the gradient of batch sample loss
l.backward()
# using small batch random gradient descent to iter model parameters
sgd([w, b], lr, batch_size)
# reset parameter gradient
w.grad.data.zero_()
b.grad.data.zero_()
train_l = loss(net(features, w, b), labels)
print('epoch %d, loss %f' % (epoch + 1, train_l.mean().item()))
输出
epoch 1, loss 0.027050
epoch 2, loss 0.000091
epoch 3, loss 0.000048
epoch 4, loss 0.000048
epoch 5, loss 0.000048
w, true_w, b, true_b
输出
(tensor([[ 1.9996],
[-3.4000]], requires_grad=True),
[2, -3.4],
tensor([4.1999], requires_grad=True),
4.2)
刚接触这些知识点,感觉代码讲解不够详细,跟起来有些吃力,后面的看了但是还没搞懂。